Autonomous vehicle with contingency consideration in trajectory realization

Patent metadata

Field	Value
Publication number	US-12570328-B2
Application number	US-202318164652-A
Country	US
Kind code	B2
Filing date	Feb 6, 2023
Priority date	Jan 27, 2023
Publication date	Mar 10, 2026
Grant date	Mar 10, 2026

Abstract

Provided are methods for determining a trajectory, which can include obtaining, using the at least one processor, sensor data associated with an environment in which a vehicle is operating, wherein the environment comprises one or more agents including a first agent; determining, using the at least one processor, based on the sensor data, a first prediction associated with the first agent; determining, using at least one processor, based on the first prediction, a primary homotopy; determining, using the at least one processor, based on the primary homotopy and the first prediction, one or more contingency homotopies associated with a contingency; determining, using the at least one processor, based on the primary homotopy and the one or more contingency homotopies, a primary trajectory; and providing, using the at least one processor, operation data associated with the primary trajectory to cause the vehicle to operate based on the primary trajectory.

First claim

What is claimed is:

1 . A method comprising:
obtaining, using at least one processor, sensor data associated with an environment in which an autonomous vehicle is operating, the environment comprising one or more agents including a first agent; determining, using the at least one processor, a first prediction associated with the first agent based at least in part on the sensor data; determining, using the at least one processor, a second prediction associated with the first agent based at least in part on the sensor data; determining, using the at least one processor, a primary homotopy associated with nominal operation of the autonomous vehicle based at least in part on the first prediction; determining, using the at least one processor, a primary trajectory located within the primary homotopy; determining, using the at least one processor, one or more contingency homotopies associated with contingent operation of the autonomous vehicle based at least in part on the second prediction; determining, using the at least one processor, one or more contingency trajectories located within the one or more contingency homotopies, wherein determining one or more contingency trajectories comprises:
determining one or more contingency costs for the one or more contingency trajectories; and
applying a set of weights to the one or more contingency costs;
modifying, using the at least one processor, the primary trajectory based at least in part on the one or more contingency homotopies, wherein the modified primary trajectory is not a contingency trajectory of the one or more contingency trajectories; and causing operation of the autonomous vehicle based at least in part on the modified primary trajectory.

2 . The method of claim 1 , wherein determining the one or more contingency homotopies based at least in part on the second prediction comprises:
adding at least one hallucinated agent in an occluded part of a map of the environment of the autonomous vehicle; and obtaining the one or more contingency homotopies from a set of predetermined worst-cases based at least in part on a specified scenario associated with the first agent proximate to the autonomous vehicle or the at least one hallucinated agent.

3 . The method of claim 2 , wherein the set of predetermined worst-cases are associated with a location of the autonomous vehicle relative to the map.

4 . The method of claim 2 , wherein the set of predetermined worst-cases are based at least in part on semantics related to the first agent and/or to the environment.

5 . The method of claim 1 , wherein determining the one or more contingency homotopies based at least in part on the second prediction comprises:
determining one or more sets of constraints based at least in part on the first prediction, wherein each set of constraints characterizes a contingency homotopy.

6 . The method of claim 5 , wherein determining the one or more sets of constraints based at least in part on the first prediction comprises:
determining the one or more sets of constraints, where at least one set of constraints of the one or more sets of constraints comprises one or more of: one or more spatio-temporal constraints and one or more station-time constraints.

7 . The method of claim 1 , wherein determining the first prediction associated with the first agent based at least in part on the sensor data comprises:
determining the first prediction associated with the first agent based at least in part on the sensor data, where the first prediction is associated with a likelihood that the first agent will perform a first action.

8 . The method of claim 1 , further comprising:
selecting a selected trajectory from among the primary trajectory and the one or more contingency trajectories, wherein causing operation of the autonomous vehicle based at least in part on the modified primary trajectory comprises:
causing operating of the autonomous vehicle based at least in part on the selected trajectory.

9 . The method of claim 1 , wherein modifying the primary trajectory based at least in part on the one or more contingency homotopies comprises:
determining the primary trajectory and one or more contingency trajectories based at least in part on one or more actions for the autonomous vehicle, wherein the modified primary trajectory and the one or more contingency trajectories share one or more actions.

10 . The method of claim 7 , wherein determining the primary trajectory within the primary homotopy and the one or more contingency trajectories within the one or more contingency homotopies comprises:
determining a nominal cost for the primary trajectory and one or more contingency costs for the one or more contingency trajectories, and applying a set of weights to the nominal cost and the one or more contingency costs.

11 . The method of claim 1 , wherein each weight of the set of weights is indicative of a probability.

12 . The method of claim 10 , wherein the set of weights comprises one or more predetermined weights.

13 . A system, comprising at least one processor; and at least one memory storing instructions thereon that, when executed by the at least one processor, cause the at least one processor to perform operations including:
obtaining, using the at least one processor, sensor data associated with an environment in which an autonomous vehicle is operating, the environment comprising one or more agents including a first agent; determining, using the at least one processor, a first prediction associated with the first agent based at least in part on the sensor data; determining, using the at least one processor, a second prediction associated with the first agent based at least in part on the sensor data; determining, using the at least one processor, a primary homotopy associated with nominal operation of the autonomous vehicle based at least in part on the first prediction; determining, using the at least one processor, a primary trajectory located within the primary homotopy; determining, using the at least one processor, one or more contingency homotopies associated with a contingent operation of the autonomous vehicle based at least in part on the second prediction; determining, using the at least one processor, one or more contingency trajectories located within the one or more contingency homotopies, wherein determining one or more contingency trajectories comprises:
determining one or more contingency costs for the one or more contingency trajectories; and
applying a set of weights to the one or more contingency costs;
modifying, using the at least one processor, the primary trajectory based at least in part on the one or more contingency homotopies, wherein the modified primary trajectory is not a contingency trajectory of the one or more contingency trajectories; and causing, using the at least one processor, operation of the autonomous vehicle based at least in part on the modified primary trajectory.

14 . The system of claim 13 , wherein determining the one or more contingency homotopies based at least in part on the second prediction comprises:
adding at least one hallucinated agent in an occluded part of a map of the environment in which the autonomous vehicle is operating; and obtaining the one or more contingency homotopies from a set of predetermined worst-cases based at least in part on a specified scenario associated with the first agent proximate to the autonomous vehicle or the at least one hallucinated agent.

15 . The system of claim 14 , wherein the set of predetermined worst-cases are associated with a

Assignees

• Motional Ad Llc

Inventors

• Spielberg Nathan
• Jespersen Thomas Kølbæk
• Mamakoukas Georgios
• Miller Jesse
• Cserna Bence

Classifications

• B60W50/0097

  Predicting future conditions · CPC title

• B60W2556/40

  High definition maps · CPC title

• B60W2050/0022

  Gains, weighting coefficients or weighting functions · CPC title

• B60W2554/00

  Input parameters relating to objects · CPC title

• B60W60/0027 Primary

  using trajectory prediction for other traffic participants · CPC title